From: Patrick McHardy <kaber@trash.net>
To: Joerg Dorchain <joerg@dorchain.net>
Cc: netfilter@vger.kernel.org
Subject: Re: nf_conntrack_sip problem
Date: Wed, 01 Jul 2009 18:16:21 +0200 [thread overview]
Message-ID: <4A4B8BD5.6060907@trash.net> (raw)
In-Reply-To: <20090701161029.GC9285@Redstar.dorchain.net>
Joerg Dorchain wrote:
> On Wed, Jul 01, 2009 at 05:05:49PM +0200, Patrick McHardy wrote:
>>> I tried this. Actually, it makes things worse. Now Asterisk
>>> complains: [Jul 1 16:17:46] WARNING[20516]: chan_sip.c:1787
>>> __sip_xmit:
>>> sip_xmit of 0x86f8de0 (len 384) to 217.10.79.9:5060 returned -1:
>>> Operation not permitted
>>>
>>> (Trying to register with sipgate.de; registration in parallel
>>> with tel.lu seems to work)
>> sipgate needs sip_direct_media=0 since the RTP streams originate from
>> a seperate cluster.
>
> I loaded the module with sip_direct_signalling=0 and
> sip_direct_media=0 to get these messages.
That should be fine. Possibly related to NAT without the helper,
see below.
>> Did you load the NAT module before the conntrack module?
>
> I did not load the nat modules at all. As said, I am only
> interested in dynamically accepting the rtp streams.
>>> nf_conntrack_sip without options on a trial incoming call however gives:
>>>
>>> # conntrack -E expect
>>> 180 proto=17 src=85.93.219.114 dst=212.88.133.153 sport=0 dport=7070
>>> 180 proto=17 src=85.93.219.114 dst=212.88.133.153 sport=0 dport=7071
>
> Also for tel.lu the expected IP should be 85.93.219.122.
It takes the addresses from the SDP payload. The gateway seems to be
handling RTP stream with a different server as well.
> BTW, it seems that combining an SER for the handling the sip part
> with an asterisk for the dial-in part seems to be common. Here it
> means the RTP stream is coming typically from a different IP than
> the register endpoint.
Thats what the sip_direct_media=0 option is meant for.
>> Besides the direct_media option, I assume you're accepting EXPECTED
>> and RELATED packets?
>
> No, only RELATED. I repeat the line: -A checkblock -m state
> --state RELATED,ESTABLISHED -j RETURN
Sorry, my mistake :) That should be fine.
> For reference, again the complete iptables:
> # Generated by iptables-save v1.4.3.2 on Wed Jul 1 13:26:32 2009
> *nat
> :PREROUTING ACCEPT [1385:93589]
> :POSTROUTING ACCEPT [319:26979]
> :OUTPUT ACCEPT [5114:401834]
> -A PREROUTING ! -i ppp0 -p udp -m udp --dport 5060 -j REDIRECT=20
> -A POSTROUTING -o ppp0 -j MASQUERADE=20
> COMMIT
So you are using NAT? You *need* the NAT helper in that case for
remapping clashing ports.
next prev parent reply other threads:[~2009-07-01 16:16 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-07-01 11:37 nf_conntrack_sip problem Joerg Dorchain
2009-07-01 12:03 ` Patrick McHardy
2009-07-01 14:43 ` Joerg Dorchain
2009-07-01 15:05 ` Patrick McHardy
2009-07-01 16:10 ` Joerg Dorchain
2009-07-01 16:16 ` Patrick McHardy [this message]
2009-07-01 20:56 ` Joerg Dorchain
2009-07-02 8:17 ` Joerg Dorchain
2009-07-02 9:04 ` Joerg Dorchain
2009-07-03 9:45 ` Patrick McHardy
2009-07-03 11:20 ` Joerg Dorchain
2009-07-03 9:44 ` Patrick McHardy
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4A4B8BD5.6060907@trash.net \
--to=kaber@trash.net \
--cc=joerg@dorchain.net \
--cc=netfilter@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.